Morphology and characteristics of laser-induced aluminum plasma in argon and in air: A comparative study

Morphology and characteristics of laser-induced aluminum plasma in argon and in air: A comparative study

    Morphology and characteristics of laser-induced aluminum plasma in argon and in air: a comparative study Xueshi Bai, Fan Cao, Vincent...

879KB Sizes 0 Downloads 53 Views

    Morphology and characteristics of laser-induced aluminum plasma in argon and in air: a comparative study Xueshi Bai, Fan Cao, Vincent Motto-Ros, Qianli Ma, Yanping Chen, Jin Yu PII: DOI: Reference:

S0584-8547(15)00236-0 doi: 10.1016/j.sab.2015.09.023 SAB 4983

To appear in:

Spectrochimica Acta Part B: Atomic Spectroscopy

Received date: Accepted date:

28 January 2015 25 September 2015

Please cite this article as: Xueshi Bai, Fan Cao, Vincent Motto-Ros, Qianli Ma, Yanping Chen, Jin Yu, Morphology and characteristics of laser-induced aluminum plasma in argon and in air: a comparative study, Spectrochimica Acta Part B: Atomic Spectroscopy (2015), doi: 10.1016/j.sab.2015.09.023

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Morphology and characteristics of laser-induced aluminum plasma in argon and in air: a comparative study

Institut Lumière Matière, UMR 5306 Université Lyon 1-CNRS, Université de Lyon,

IP

a

T

Xueshi Baia, Fan Caoc, Vincent Motto-Rosa, Qianli Mab, Yanping Chenb and Jin Yua,b,

b

SC R

69622 Villeurbanne CEDEX, France

Key Laboratory for Laser Plasmas (Ministry of Education), Department of Physics and Astronomy, Shanghai Jiao Tong University, Shanghai 200240, China School of Physics and Technology, Wuhan University, Wuhan, 430072, PR China

NU

c

MA

Abstract

In laser-induced breakdown spectroscopy (LIBS), ablation takes place in general in an

D

ambient gas of the atmospheric pressure, often in air but also in noble gas such as argon or

TE

helium. The use of noble gas is known to significantly improve the performance of the technique. We investigate in this work the morphology and the characteristics of induced

CE P

plasma in argon and in air. The purpose is to understand the mechanism of the analytical performance improvement by the use of argon ambient with respective to air ambient and

AC

the dependence on the other experimental parameters such as the laser fluence. The observation of plasma morphology in different ambient gases provides also information for better design of the detection system which optimizes the signal collection according to the used ambient gases. More specifically, the expansion of the plasma induced on an aluminum target with nanosecond infrared (1064 nm) laser pulse in two ambient gases, argon and the atmospheric air, has been studied with spectroscopic imaging at short delays and with emission spectroscopy at longer delays. With relatively low ablation laser fluence (65 J/cm2), similar morphologies have been observed in argon and in air over the early stage of plasma expansion, while diagnostics at longer delay shows stronger



Corresponding author. E-mail address: [email protected] (Jin Yu). 1

ACCEPTED MANUSCRIPT emission, higher electron density and temperature for plasma induced in argon. With higher ablation laser fluence (160 J/cm2) however, different expansion behaviors have been observed, with a stagnating aluminum vapor near the target surface in air while a

T

propagating plume away from the target in argon. The craters left on the target surface

IP

show as well corresponding difference: in air, the crater is very shallow with a target

SC R

surface chaotically affected by the laser pulse, indicating an effective re-deposition of the ablated material back to the crater; while in Ar a deeper crater is observed, indicating an efficient mass removal by laser ablation. At longer delays, a brighter, denser and hotter

NU

plasma is always observed in argon than in air as with lower ablation laser fluences. The

MA

observed different influences of the ambient gas on the plasma expansion behavior for different laser fluences are related to the different modes of laser-supported absorption waves, namely laser-supported combustion (LSC) wave and laser-supported detonation

TE

D

(LSD) wave.

CE P

Keywords: Laser-induced plasma, Ambient gas, Laser-induced absorption waves, Plume morphology, Ablation crater

AC

1. Introduction

Laser-induced breakdown spectroscopy (LIBS) uses ablation plasma as a spectroscopic emission source. The analytical performance of the technique greatly depends on the properties of the plasma and the corresponding properly designed optical systems to generate the plasma and to collect the emission from the plasma. Even working under the fixed atmospheric pressure as in most of LIBS measurements, it is well known that the nature of the ambient gas (molecular or monoatomic gas, reactive or inert gas) can greatly influence the expansion behavior of the plasma and the properties of the resulted plume [1-5], thus crucially affect the analytical performance of the technique. It has been clearly established that a noble gas such as argon can lead to a 2

ACCEPTED MANUSCRIPT dense and hot plasma and therefore stronger spectral emission from the plasma than that in the case with the atmospheric air as ambient gas. The reported results however are often based on time-resolved and space-integrated diagnostics. Observation on the

T

morphology of the plasma, and especially over the early stage of the plasma expansion, is

IP

often absent. Such absence, of cause, does not alternate the established results about the

SC R

beneficial effects of the use of argon for a better LIBS performance. But it prevents from a thorough understanding of the mechanism responsible of such improvement. In addition, the morphological information of the plume, including its size, the distribution

NU

of the species and its temporal evolution, is very important for the design of a proper

MA

detection system of an optimized collection of the plasma emission. The understanding and the control of the morphology of the plasma are therefore important for the analytical performance of LIBS in terms of precision, repeatability and reproducibility [6].

D

According to the results presented in our previous works [7-11], laser-supported

TE

absorption waves (LSAW) during the post-ablation interaction, especially the

CE P

laser-supported consumption (LSC) wave and laser-supported detonation (LSD) wave can greatly influence and modify the morphology of the plasma. Such influence is particularly pronounced for infrared (IR, 1064 nm) ns laser ablation at relatively high

AC

fluence (> 150 J/cm2). However up to now, the influence of the nature of the ambient gas (monoatomic or molecular) on the expansion behavior of the induced plasma remains still unclear.

In this paper, we report the results of a comparative study of the morphology and the characteristics of laser induced plasmas in different ambient gases of argon and air at one atmosphere pressure. A simple configuration of ablation with an IR (1064 nm) nanosecond pulse of a metallic target (aluminum) was used in order to stress the behavior of the plasma during its expansion into different ambient gases. The expansion process of plasma was studied in two fluence regimes, a moderate fluence regime of 65 J/cm2 and a high fluence regime of 160 J/cm2. In each of these ablation regimes, the expansion 3

ACCEPTED MANUSCRIPT behavior of the plasma was observed using fast spectroscopic imaging at short delays (0 to 200 ns) and space-resolved emission spectroscopy at longer delays (350 to 2000 ns). Fast spectroscopic imaging reveals the morphology of the plasma [12], while emission

T

spectroscopy results in electron density and temperature profiles of the plasma. The

IP

observation of the plasma morphology in the early stage of the expansion and that of the

SC R

plasma characteristics over the delay interval when LIBS measurements generally take place, are further correlated to the observation of the craters left on the sample surface with scanning electronic microscope (SEM). Such multi-time scale and multi-aspect

NU

observations enrich the obtained results for an advanced understanding of the implicated

MA

mechanisms in different ablation fluence regimes and with different ambient gases. We interpret the observed plasma expansion behaviors and the resulting plasma characteristics using the different modes of LSAW in the early stage of the plume

D

expansion when the laser pulse is still present.

TE

2. Experimental setups and measurement protocols

CE P

Detailed description of the used experimental setup can be found elsewhere [9, 10]. A Nd:YAG laser (Quantel Brilliant) was used for ablation in two fluence regimes with pulse energy of 20 mJ corresponding to a fluence of 65 J/cm2 (moderate fluence regime)

AC

and 50 mJ corresponding to a fluence of 160 J/cm2 (high fluence regime). The aluminum targets used in the experiment were of two different qualities, which respectively correspond to two types of detection, the image type and the spectroscopy type, performed in the experiment. A piece of pure aluminum (Al 99.99%, Cu 0.005%, Si 0.002%, Fe 0.001%) was used for the image type measurement, while a certified reference aluminum alloy (Al 89.5%, Si 8.39%, Fe 0.999% and some traces) was used for the measurement of spectroscopy type. A pair of tubes installed above the target close to the laser ablation zone was used to deliver a stream of argon gas of a fixed flow of 8 /min, which ensured the plasma to expand into a pure argon ambient gas at the atmospheric pressure. The tubes were removed for ablation in the atmospheric air. 4

ACCEPTED MANUSCRIPT In the image type measurement, a pair of achromatic lens (in BK7) was used to form plasma image on an ICCD camera (Andor Technology, iStar) along an axis perpendicular to the propagation axis of the laser beam which was focused vertical

T

down to the surface of the target. Fast spectroscopic images were realized

IP

correspondingly for the different species in the plume with the help of narrowband filters

SC R

with central wavelength and corresponding species as shown in Table 1. We can see in the table that a plume in our experiment is characterized by 4 species: neutral aluminum (Al I), ionized aluminum (Al II), neutral gas (Ar I or N I according to the used ambient

NU

gas), and ionized gas (Ar II or N II). For a given species in the plume, a pair of filters was

MA

used with one of them centered on the corresponding emission line (filter-on) and the other shifted outside but nearby the line (filter-off). Assuming that the spectral intensity of the continuum background remains almost constant in the vicinity of the considered

D

emission line, it is possible to subtract the contribution of the continuum from the

TE

intensity recorded with the filter-on using the intensity recorded using the corresponding

CE P

filter-off. After such subtraction between the images, the resulting image corresponds to the emission image of the concerned species. The Abel-inversion [13] is further applied to emission image, resulting in emissivity image of the corresponding species. In our

AC

experiment, the intensity distribution in an emissivity image is considered as representative of the space distribution of the corresponding species. This consideration obviously corresponds to an approximation, because it is well known that the emissivity of an emission line is directly related to the population of the concerned species in the up state of the corresponding transition. Such approximation is however justified when the variation of the electron temperature within the plasma remains relatively smooth. This condition will be verified in our experiment. For the image recording, each raw emission image, with filter-on or with filter-off, was the result of the accumulation of 100 laser impacts distributed over 10 craters with 10 impacts by crater.

5

ACCEPTED MANUSCRIPT Table 1 Emission lines chosen to represent different species in the plume and the narrowband filters used in the image type measurement to perform spectroscopic image of these species.

Neutral (Al I)

394.4, 396.2

Ion (Al II)

358.7

Neutral (Ar I)

750.4, 751.5

on : 400

Neutral (N I)

D

Air (N)

on : 360 off : 380 on : 750 off : 720 on : 488 off : 530 on : 750

746.8 off : 720 on : 500

500.1, 500.7 off : 530

CE P

TE

Ion (N II)

off : 380

484.8, 488.0

MA

Ion (Ar II)

NU

Al

Ar

Central wavelength of the filters (nm)

T

Emission line (nm)

IP

Species

SC R

Element

In the spectroscopy type measurement, a pair of fused silica lens was used to form an image of the plasma along an axis perpendicular to the laser beam incident direction. An

AC

optical fiber of aperture of 50 µm was put in the plasma image plane to catch the emission from a given volume of the plasma. Axial profile of the plasma was thus preformed in the lateral middle of the plasma image by translating the fiber step by step along the laser incidence axis. Such detection system allowed a space resolution of 75 µm by taking into account the magnification of the used optical system. The output of the fiber was connected to the entrance of an echelle spectrometer, which was in turn connected to an ICCD camera (Mechelle 5000 and iStar from Andor Technology). Each spectrum was the result of the accumulation over 200 laser impacts distributed over 20 craters with 10 impacts by crater.

6

ACCEPTED MANUSCRIPT Representative lines were chosen in order to get axial profiles of the ablated aluminum plume. For aluminum ion, the line at 281.6 nm was chosen with the lower energy level of the transition of 11.8 eV, which prevents it from significant

T

self-absorption. For aluminum atom, the resonant line at 309.3 nm (with the ground state

IP

as the lower level of the transition) was among the several lines that exhibited an enough

SC R

good signal-to-noise ratio, because of the high degree of ionization of the aluminum vapor in the condition of our experiment, especially at short delays. Such high ionization reduced the self-absorption of the chosen neutral aluminum line. The electronic density

NU

was determined using the Stark broadening of the Ar I 696.5 nm line in the argon ambient,

MA

and that of the H I 656.2 nm line in the air ambient. The determination of the electron temperature needs the plasma to be in the local thermodynamic equilibrium (LTE). Our previous work shows that in the delay interval considered for the spectroscopy type

D

measurement (350 to 2000 ns), the LTE state represents a reasonably good approximation

TE

for the studied plasma. [14] Multi-element Saha-Boltzmann plot [15] was thus used to

CE P

deduce the electron temperature. We estimate the relative standard deviations of the determination of the electron density and the electron temperature to be 15% and 10% respectively.

AC

3. Results and discussions 3.1. Moderate fluence ablation regime 3.1.1. Emissivity images in the early stage of the plasma expansion With an ablation laser energy of 20 mJ (corresponding to a fluence of 65 J/cm2), we have first observed the expansion behavior of the plume in the two types of ambient gas in the early stage of the expansion from the initiation of the plasma to about 200 ns. Notice that the internal delay of the used ICCD camera did not allow an observation delay shorter than 40 ns. Images obtained with the superposition of the emissivity images corresponding to the 4 representative species of the plume are shown in Fig. 1 for ablations in the two types of ambient gas. Notice that the images in the left column 7

ACCEPTED MANUSCRIPT correspond to the plasma induced in the argon ambient, while those in the right column to that induced in the air ambient. Notice also that the frame of the shown images has a dimension of 1.5 mm

1.5 mm, and that the bottom line of the images represents the

T

target surface. Let us look at now the details in these images. The emissivity images of the

IP

various species are displayed in different colors with blue for neutral gas (Ar I or N I),

SC R

gray for ionized gas (Ar II or N II), green for neutral aluminum (Al I) and red for ionized aluminum (Al II). The emissivity of each species is normalized to its own maximum in order to better show the internal structure of a plasma. Along the axial direction of the

NU

plume, a layered structure of emissivity can be observed for the plume induced in argon

MA

as well as for that induced in air with a similar arrangement in the both cases. If we accept here the representativity of the emissivity distribution for the corresponding species distribution under the condition mentioned in the section 2 above, the emissivity images

D

in Fig. 1 show actually a layered structure of the different species in the plume. Starting

TE

from the target surface and going upward toward the incidence direction of the laser

CE P

beam, one can successively observe the layer of neutral aluminum, the layer of ionized aluminum, the layer of ionized gas, argon or nitrogen, and finally the layer of neutral gas, argon or nitrogen. Notice the absence of the neutral aluminum layer for ablation in the air

AC

ambient at short delays.

8

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

Fig. 1. Time-resolved emissivity images of plasma induced in argon (the left column) and in air (the right column) with laser pulse energy of 20 mJ and corresponding fluence of 65 J/cm2. The detection delay of each image is indicated on the image. The real dimension of each picture is 1.5 mm

1.5 mm, and the bottom line of the images represents the target surface. False colors

9

ACCEPTED MANUSCRIPT are used to represent the different species in the plume with blue for neutral argon or neutral nitrogen, grey for ionized argon or ionized nitrogen, red for ionized aluminum, and green for

IP

T

neutral aluminum. The emissivity of each species is normalized to its own maximum.

SC R

Such arrangement of species in successive layers within a nearly spherical plasma morphology corresponds to the consequence of LSC wave as we discussed in our previous works [10, 11]. In the LSC wave model, the layer of shocked ambient gas

NU

remains transparent for the ablation laser radiation during the post ablation interaction. It is why after the initiations of the ablation plume and the consequent shockwave, the

MA

tailing part of the ablation pulse transmits through the shocked gas layer and touches the ablation vapor, aluminum vapor for instance. Absorption of laser energy occurs during

D

the propagation of the laser pulse in the vapor mainly through the inverse bremsstrahlung,

TE

which in turn leads to further ionization of the vapor through avalanche process. Consequently, two types of layered structure can be resulted from this interaction

CE P

configuration. i) If the ablation vapor is enough dense and the laser pulse is not enough strong after its transmission through the layer of shocked gas, the vapor can only be

AC

ionized in its top part first touched by the laser pulse. In the lower part of the vapor near the target surface, neutral aluminum can be observed. This corresponds to the case of the argon ambient (Fig. 1, left column). ii) If the ablation vapor is not enough dense and the laser pulse is enough strong after its transmission through the layer of shocked gas, the vapor can be totally ionized. This corresponds to the case of the air ambient (Fig.1, right column). At the same time, the fact that the major part of the ablation laser pulse transmits through the shocked ambient gas layer implies that the ionization of the ambient gas due to absorption of the laser pulse is negligible. However a population of ionized ambient gas is clearly observed Fig. 1 for plasmas induced in argon and in air with a location under the layer of excited neutral gas and above the ionized aluminum. The location of the 10

ACCEPTED MANUSCRIPT ionized gas behind a significant layer of excited neutral gas at short delays, as can be observed in Fig. 1, is not compatible with a direct ionization of the gas by the laser pulse, because in this case the ionization front should face to the laser pulse propagation

T

direction, i.e. locating above the excited neutral gas, at least at very short delay just at the

IP

end of the laser-supported absorption wave. It is therefore reasonable to deduce that the

SC R

observed ionized gas population is due to the interaction with the ablation vapor which is heated by the tailing part of the ablation laser pulse to high temperature. Such interaction can be collisional excitation and ionization and/or absorption of the UV radiation emitted

NU

by the vapor. [16] The layer of the excited shocked gas close to the ablation vapor can

MA

thus be ionized through these interactions. The formation of the ionized ambient gas above the ablation vapor would be thus ulterior to the transmission of the tailing part of the laser pulse through the shocked gas layer, in such way this does not contribute to the

D

shielding of the laser pulse. The plasma shielding of the tailing part of the ablation laser

TE

pulse is therefore totally due to the ablation vapor. Such situation corresponds well to the

CE P

LSC mode of laser-supported absorption wave as mentioned above. Plasmas induced with modest ablation laser fluence in argon and in air show therefore similar morphologies. More generally, we can say that when the LSC wave represents the

AC

dominate mechanism of the post ablation interaction, the nature of the ambient gas does not influence greatly the propagation behavior of the plasma. Qualitatively this can be easily understood since in the LSC wave propagation, the layer of shocked ambient gas, regardless whether it is argon or the atmospheric air, remains transparent for the ablation laser radiation. Thus the different optical properties of the ambient gases are not involved in the determination of the propagation behavior of the plasma. Some minor differences observed between the plasma morphologies in the two kinds of ambient gas may be rather related to the hydrodynamic or thermodynamic properties of the gases. For example, air is more compressible than argon (

= 1.4018,

= 1.6696) [17]. This leads to a larger

global extent of the plume at a given delay in air than in argon as shown in Fig. 1. 11

ACCEPTED MANUSCRIPT 3.1.2. Axial profile of the plasma at longer delays Beyond 300 ns, the continuum emission, very strong in the early stage of the plasma expansion, becomes damped to an enough weak level for the emission spectroscopy to be

T

efficient for plasma diagnostics. Such diagnostics provides quantitative information

IP

about the plasma in its evolution with the axial profiles of the species, the electron density

SC R

and temperature. This detailed information about the characteristic parameters of the plasma completes the observation on the morphology of the plasma in short delay presented in the above section. Especially, the temperature profile, even though measured

NU

after the interval of emissivity image recording, can show the variation of the temperature

MA

in the plume and thus provide the justification for the representativity of the emissivity distribution for the species distribution within the plasma. The obtained results are shown in Fig. 2. Notice that the abscissa of the figures

D

represents the axial distance from the target with the origin of the axis corresponding to

TE

the surface. Notice also that the emission intensities were recorded from the plasmas

CE P

induced in argon and in air in the same condition in such way that they are directly comparable. We can first remark that the emission intensities of neutral and ionized aluminum from the plasma induced in argon [Fig. 2 (a) and (b)] are much stronger than

AC

those from the plasma induced in air [Fig. 2 (c) and (d)]. This corresponds to the benefic effect of argon in LIBS measurements reported in the literature and mentioned in the introduction of the paper. The enhancement in our case is about of a factor of 6 for neutral and 12 for ion as shown in Fig. 2 (a to d). Such enhancement of the emission intensities is consistent with the larger electron density and temperature for the plasma induced in argon comparing to those of the plasma induced in air as shown in Fig. 2 (e) and (f). Let us look at now the form of the profiles of aluminum vapor in argon and in air. We can see that the aluminum vapor in air presents a larger extent than that in argon, which is consistent with the larger compressibility of the air mentioned above. We can see also that the axial extent of electron density (Fig. 2 e) corresponds well to the axial profile of 12

ACCEPTED MANUSCRIPT aluminum ions [Fig. 2 (b) and (d)], which confirms the fact that the electron density is mainly contributed by ionization of aluminum vapor. Finally we can remark in Fig. 2 (f) that the variation of the temperature across axial direction of the plasma is quite smooth,

T

the total variation is less than 10%. Such smooth temperature variation justifies the

IP

representativity of the emissivity distribution for the species distribution within the

SC R

plasma. Our interpretation of the Fig. 1 and Fig. 2 (a-d) in terms of the different species in

AC

CE P

TE

D

MA

NU

the plasma can be therefore justified.

13

ACCEPTED MANUSCRIPT Fig. 2. Profiles of the plasma along the axis of the laser incident direction and in the lateral middle of the plasma. (a) and (b): respectively neutral and ionized aluminum in argon ambient; (c) and (d): respectively neutral and ionized aluminum in air ambient; (e): electron density in

T

argon and air; (f): electron temperature in argon and air. The detection delays are indicated in

IP

the figures. Ablation laser energy was 20 mJ per pulse and corresponding fluence 65 J/cm2. The

SC R

symbols used to specify the ambient gas and the detection delay are explained in (a) to (b) and are also used for (e) and (f).

NU

In a practical point of view for LIBS measurements, we can say that with moderate

MA

ablation fluence, only small difference exists between the morphologies of the plasmas induced in argon and in air. The emission intensity is however much higher from the plasma induced in argon than that induced in air with same laser fluence. Such signal

D

enhancement is known in the LIBS literature and corresponds to the benefice of the use of

TE

argon as background gas. Two effects can contribute to such signal enhancement. The

CE P

first is clearly a higher temperature of the plasma induced in argon due to a smaller thermal conductivity of argon (16.48 mWm-1K-1) with respect to air (24.36 Wm-1K-1) [17]. The second effect corresponds to a better confinement, thus a smaller extent, of the

AC

plasma induced in argon due to a smaller compressibility of argon as mentioned above.

3.2. High fluence ablation regime 3.2.1. Emissivity images in the early stage of the plasma expansion With an ablation laser energy of 50 mJ (corresponding to a fluence of 160 J/cm2), we have observed the expansion behavior of the plume in the two types of ambient gas in the early stage of the expansion from the initiation of the plasma to a delay of about 200 ns. The detection condition and the method used to display the obtained emissivity images remain the same as those used in section 3.1.1 and the results are shown in Fig. 3. Notice that the images in the left column correspond to the plasma induced in the argon ambient, while those in the right column to that induced in the air ambient. Notice also that the 14

ACCEPTED MANUSCRIPT frame of the shown images has a dimension of 1.5 mm

1.5 mm, and that the bottom

line of the images represents the target surface. Here again we assume the representativity of the emissivity distribution for the corresponding species distribution within the plume

T

as in the section 3.1.1. Let us look at now the details in these images. For the both ambient

IP

gases, a population of ionized ambient gas can be observed with a large spatial extent

SC R

which occupies almost the entire volume of the plasma. In contrast, population of the excited neutral ambient gas is almost absent at very short delay. This situation corresponds to a nearly complete ionization of the shocked ambient gas. As we discussed

NU

in our previous works, such important ionization of the ambient gas over an extended

MA

zone in the front of the shocked gas is the result of the ignition of laser-supported detonation wave (LDS) [10,11]. Our previous works further demonstrated that the ignition of LSD for an ablation plume propagating in an argon ambient gas leads to a

D

mixture of ions of argon and the ablation vapor, aluminum for instance, within a large

TE

zone which consists in the core part of the plume [11]. This phenomenon is again

CE P

observed in the present experiment as shown in the image in the left column in Fig. 3. At longer delays, neutral ambient gas appears first in the lower peripheral of the plasma in

AC

argon and on the top of the plasma in air.

15

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

Fig. 3. Time-resolved emissivity images of the plasma induced in argon (the left column) and in air (right column) with laser pulse energy of 50 mJ and laser fluence of 160 J/cm2. The detection delay of each image is indicated in the image. The real dimension of each image is 1.5 mm

16

ACCEPTED MANUSCRIPT 1.5 mm, and the bottom line of the images represents the target surface. False colors are used to represent the different species in the plume with blue for neutral argon or neutral nitrogen, grey for ionized argon or ionized nitrogen, red for ionized aluminum, and green for neutral

IP

T

aluminum. The emissivity of each species is normalized to its own maximum.

SC R

It is however very striking to remark that for the plasma induced in air, the above mentioned mixing between the ions of ambient gas and aluminum vapor observed for the plasma induced in argon is totally absent. On the contrary, the population of aluminum

NU

ions is observed to be confined in a small volume in the lower part of the plume near the

MA

target surface. And such stagnating population of aluminum ion presents a very small axial extent compared to that observed in argon. Thus our results clearly show that for high laser fluence ablation when LSD is ignited, the two types of ambient gas lead to very

D

different plasma morphologies, a largely extended and propagating vapor plume for

TE

ablation in argon and a much less extended and stagnating vapor plume for ablation in air.

CE P

In addition, the vapor plume is totally mixed with ionized argon, while in air the vapor plume is confined in a volume close to the target surface and separated from the ionized shocked gas (N II). We can also remark in Fig. 3 that the axial extent of the plume induced

AC

in argon is larger than that induced in air (a contrary situation with respect to modest ablation laser fluence). This result can be the consequence of a faster propagation of the LSD wave in argon than in air as indicated by [18] (1) where

is the velocity of the LSD wave,

intensity and

the ratio of specific heats,

density of the ambient gas. Using the values of

laser

, provided in the

section 3.1.1 above for argon and air and with the density of argon and air in the standard condition, we can determine the relation between the LSD velocities in argon and in air: (2). 17

ACCEPTED MANUSCRIPT A final remark in this section concerns the observed difference about the mixing or the separation between the ionized ambient gas and the ablated aluminum vapor. We think that a thorough discussion for understanding the involved mechanisms would

T

lengthen too much the paper. Such attempt is out of scope of this paper mainly intending

IP

to expose the observed experimental evidences. We can instead just evoke here some

SC R

ideas for a full investigation. We believe that such difference in the behaviors of the LSD in argon and in air comes from the different ionization mechanisms for these 2 gases. For argon, an atomic gas, the mechanism of ionization seems quite clear. It should involve

NU

collisional electronic excitations in the shockwave followed by photoionizations from the

MA

excited states of atomic argon. So this is a pure electronic process and should have a fast time constant. While for air, molecular processes should be involved in its ionization. Such processes may include the dissociation of the air molecules O2 and N2 in the

D

shockwave, followed by associative ionizations [19]. It is therefore quite obvious that in

TE

air, the involved molecular processes in its ionization lead to a much longer time constant

CE P

for the ignition of the LSD than in an atomic gas like argon. The different time constants for the LSD wave ignition should be therefore the origin of the observed different behaviors of the LSD wave in air and in argon.

AC

3.2.2. Axial profile of the plasma at longer delays As in the section 3.1.2, the axial profile of the plasma was performed with high ablation laser fluence. The obtained results are shown in Fig. 4. As in Fig. 2, the abscissa of the figures represents the distance from the target with the origin of the axis corresponding to the surface. Also as is Fig. 2, the emission intensities were recorded from plasma induced in argon and in air in the same condition in such way that they are directly comparable. Similarly for the intensity profiles shown in Fig. 2, much higher emission intensity is observed for the plasma induced in argon than in air, for neutral aluminum [Fig. 4 (a) and (c), ~ 10 times higher] as well as ionized aluminum [Fig. 4 (b) and (d), ~ 20 times higher]. In addition, the propagation behaviors can be observed to be 18

ACCEPTED MANUSCRIPT very different for plasmas induced in argon and in air. In argon, the aluminum plume exhibits a propagation away from the target surface over the period of observation from 500 ns to 2000 ns. Such propagation concerns neutral aluminum [Fig. 4 (a)] as well as

T

ionized aluminum [Fig. 4 (b)]. In air on the contrary, the aluminum vapor mainly remains

IP

static close to the target surface (0 mm to ~ 1.0 mm) without significant propagation of

SC R

the plume away from the target surface. This observation is coherent with respective to the imaging observation shown in Fig. 3. Let us look at now Fig. 4 (e) and (f) for the axial profiles of the electron density and temperature. Higher electron density and temperature

NU

are observed for plasma induced in argon with respective to that induced in air. Here

MA

again different propagation behaviors can be observed between plasmas induced in argon and air. In argon, both the profiles of electron density and temperature exhibit a propagation away from the target surface, while in air the both profiles remain relative

D

static and close to the target surface. Finally the relatively smooth variation of the electron

TE

temperatures shown in Fig. 4 (f) over the main extent of the plasmas, again justifies the

AC

plume.

CE P

representativity of the emissivity distribution for the species distribution within the

19

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

Fig. 4. Profiles of the plasma along the axis of the laser incident direction and in the lateral middle of the plasma. (a) and (b): respectively neutral and ionized aluminum in argon ambient; (c) and (d): respectively neutral and ionized aluminum in air ambient; (e): electron density in argon and air; (f): electron temperature in argon and air. The detection delays are indicated in the figures. Ablation laser energy 50 mJ per pulse and corresponding fluence 160 J/cm2.

In the practical point of view for LIBS applications, when ablation laser fluence becomes quite high, the emission zone from the ablation plume can be quite different for 20

ACCEPTED MANUSCRIPT the plume induced in argon and for that induced in air. In air, such zone is confined quite close to the target surface in contrast with a larger emission zone propagating away from the target surface in argon. In the case where the plasma emission is detected along an

T

axis perpendicular to the laser incidence axis, a proper design of the detection system

IP

needs to be optimized for each of these two situations. Similar to the modest ablation

SC R

fluence, the emission is always enhanced with the argon ambient, confirming thus the general conclusion about the interest of argon for LIBS measurements in the literature.

NU

3.3. Craters produced in different ambient gases with moderate and high ablation laser fluences

MA

In order to correlate the plasma expansion behavior to the morphology of the corresponding ablation crater left on the target surface with different ablation fluences

D

and in different ambient gases, the craters were observed with scanning electronic

TE

microscope. Figure 5 shows the obtained results with 3 successive ablations for each crater. We can see that with the moderate ablation fluence of 65 J/cm2, the morphologies

CE P

of the craters produced in argon and air are quite similar as shown in Fig. 5 (a) and (b), in the sense that we can observe a well formed crater indicating a similar and effective mass

AC

removal by ablation in the both cases. This confirms well our results and discussions in the section 3.1.1 where similar expansion behaviors are observed for plasma induced in argon and in air. For the high ablation fluence at 160 J/cm2, in argon [Fig. 5(c)], a relatively deep crater is observed, showing efficient mass removal by ablation. On the contrary in air [Fig. 5 (d)], the laser impact is only distinguishable with irregular surface modification. Such surface morphology suggests a significant material re-deposition after being first removed from the surface. The surface level in the crater can even be higher than the unaffected surface. Such difference observed between the craters left by ablation in argon and in air corresponds well to the difference observed for the expansion behaviors of the plasmas induced in these two different ambient gases. In argon, a propagating ablation vapor facilitates the mass removal, while in air a stagnating ablation 21

ACCEPTED MANUSCRIPT vapor near the target surface may lead to its re-deposition in the crater during the re-condensation of the plume. The observed crater morphology is therefore consistent with the imaging and emission spectroscopy observations presented in the above

CE P

TE

D

MA

NU

SC R

IP

T

sections.

AC

Fig. 5. SEM images of ablation craters in different experimental conditions of ablation fluence and ambient gas. (a) 65 J/cm2 ablation laser fluence in argon, (b) 65 J/cm2 ablation laser fluence in air, (c) 160 J/cm2 ablation laser fluence in argon and (d) 160 J/cm2 ablation laser fluence in air. Each crater accumulates 3 laser shots.

4. Conclusion In conclusion, we have in this paper investigated the expansion behavior of laser-induced plasma in different ambient gases and with different ablation laser fluences. The obtained results are first presented for a moderate ablation laser fluence of 65 J/cm2 (20 mJ pulse energy). Through imaging and time- and space-resolved emission 22

ACCEPTED MANUSCRIPT spectroscopy observations, we get the conclusion that the plasmas induced in the two types of ambient gas exhibit quite similar propagation behaviors with comparable morphology and internal structure. The observed morphology and structure of the

T

plasma suggest a post ablation interaction dominated by the LSC wave. Diagnostics

IP

with emission spectroscopy shows a stronger emission from the plasma induced in

SC R

argon together with higher electron density and temperature. The results obtained with a higher ablation laser fluence of 160 J/cm2 (50 mJ pulse energy) are then presented. The morphology and the structure of the plasmas induced in the both ambient gases exhibit

NU

the characteristics of a post ablation interaction dominated by the LSD wave. A

MA

significant difference in propagation behavior is observed between the plasmas induced in argon and in air. The most striking feature is that in argon, a large zone of mixing between ions of argon and the aluminum vapor is observed, as we reported in our

D

previous works. However for the plasma induced in air with the same laser fluence, a

TE

population of aluminum vapor confined in a limited volume close to the target surface

CE P

and separated from the shocked gas is observed. Diagnostics with emission spectroscopy confirms a propagating aluminum vapor away from the target in argon and a stagnating aluminum population in air. A more intense emission from the aluminum vapor together

AC

with higher electron density and temperature are always observed for the plasma in argon. Observation of ablation carters by SEM reveals similar morphologies for the caters left by ablations with modest laser fluence of 65 J/cm2 in argon and in air. However the same observation of the craters left by ablations with higher laser fluence of 160 J/cm2 shows significant difference between ablations in argon and in air. Such difference corresponds well to the observed propagating ablation plume in argon and stagnating one in air. Although the ensemble of the observations appears quite clear and coherent, the understanding of the underlining mechanisms still needs further experimental and theoretical investigations. Such understanding may especially concern the observed

23

ACCEPTED MANUSCRIPT difference between the propagation behaviors of the plasmas induced with high ablation laser fluence in the two ambient gases of argon and air.

T

Acknowledgements

IP

One of the authors (X.S.B.) thanks the China Scholarship Council (CSC) for their

SC R

support. The authors thank the French Rhone-Alps Region for their supports through the CMIRA international collaboration and exchange program.

References

S. S. Harilal, C. V. Bindhu, V. P. N. Nampoori and C. P. G. Vallabhan, Influence of

NU

[1]

MA

ambient gas on the temperature and density of laser produced carbon plasma, Appl. Phys. Lett. 72 (1998)167-169. [2]

J. A. Aguilera, C. Aragón, A comparison of the temperatures and electron densities

D

of laser-produced plasmas obtained in air, argon, and helium at atmospheric

A. Löbe, J. Vrenegor, R. Fleige, V. Sturm, R. Noll, Laser-induced ablation of a

CE P

[3]

TE

pressure, Appl. Phys. A 69 (1999) S475–S478.

steel sample in different ambient gases by use of collinear multiple laser pulses, Anal. Bioanal. Chem. 385 (2006) 326–332. N. M. Shaikh, S. Hafeez, M.A. Baig, Comparison of zinc and cadmium plasma

AC

[4]

parameters produced by laser-ablation, Spectrochim. Acta B, 62 (2007) 1311-1320. [5]

S. Bashir, N. Farid, K. Mahmood, M. S. Rafique, Influence of ambient gas and its pressure on the laser-induced breakdown spectroscopy and the surface morphology of laser-ablated Cd, Appl. Phys. A 107 (2012) 203–212.

[6]

V. Motto-Ros, E. Negre, F. Pelascini, G. Panczer, J. Yu, Precise alignment of the collection fiber assisted by real-time plasma imaging in laser-induced breakdown spectroscopy, Spectrochim. Acta B, 92 (2014) 60 – 69.

24

ACCEPTED MANUSCRIPT [7]

Q.L. Ma, V. Motto-Ros, F. Laye, J. Yu, W.Q. Lei, X.S. Bai, L.J. Zheng and H.P. Zeng, Ultraviolet versus infrared: effects of ablation laser wavelength on the expansion of laser-induced plasma into one-atmosphere argon gas, J. Appl. Phys.,

Q. L. Ma, V. Motto-Ros, W. Q. Lei, M. Boueri, X. S. Bai, L. J. Zheng, H. P. Zeng,

IP

[8]

T

111 (2012) 053301.

SC R

and J. Yu, Temporal and spatial dynamics of laser-induced aluminum plasma in argon background at atmospheric pressure: interplay with the ambient gas, Spectrochim. Acta Part B 65 (2010) 896-907.

X.S. Bai, Q.L. Ma, V. Motto-Ros and J. Yu, D. Sabourdy, L. Nguyen and A.

NU

[9]

MA

Jalocha, Convoluted effect of laser fluence and pulse duration on the property of a nanosecond laser-induced plasma into an argon ambient gas at the atmospheric pressure, J. Appl. Phys., 113 (2013) 013304.

D

[10] X.S. Bai, Q.L. Ma, V. Motto-Ros, J. Yu, D. Sabourdy, L. Nguyen and A. Jalocha,

TE

Experimental study of laser-induced plasma: influence of laser fluence and pulse

CE P

duration, Spectrochim. Acta B, 87 (2013) 27– 35. [11] Q.L. Ma, V. Motto-Ros, X.S. Bai and J. Yu, Experimental investigation of the structure and the dynamics of nanosecond laser-induced plasma in one-atmosphere

AC

argon ambient gas, Appl. Phys. Lett. 103 (2013) 204101. [12] V. Motto-Ros, Q.L. Ma, S. Grégoire, W.Q. Lei, X.C. Wang, F. Pelascini, F. Surma, V. Detalle, and J. Yu, Dual-wavelength differential spectroscopic imaging for diagnostics of laser-induced plasma, Spectrochim. Acta B 74-75 (2012) 11-17. [13] O.H. Nestor and H.N. Olsen, Numerical methods for reducing line and surface probe data, SIAM Rev. 2 (1960) 200-207. [14] W.Q. Lei, V. Motto-Ros, M. Boueri, Q.L. Ma, D.C. Zhang, L.J. Zheng, H. P. Zeng, and J. Yu, Time-resolved characterization of laser-induced plasma from fresh vegetables, Spectrochim. Acta Part B 64 (2009) 891-898.

25

ACCEPTED MANUSCRIPT [15] J.A. Aguilera, C. Aragón, Multi-element Saha–Boltzmann and Boltzmann plots in laser-induced plasmas, Spectrochim. Acta Part B 62 (2007) 378-385. [16] R.D. Sacks and J.P. Walters, Short-time, spatially-resolved radiation processes in

T

a high-voltage spark discharge, Anal. Chem. 42 (1970) 61-84.

IP

[17] http://encyclopedia.airliquide.com

SC R

[18] R. G. Root, Ch. 2: Modeling of Post-Breakdown Phenomena in Laser-Induced Plasmas and Applications, ed. L. J. Radziemski and D. A. Cremers, New York: Dekker 1989.

NU

[19] Ya. B. Zel’dovich and Yu. P. Raizer, in Physics of Shock Waves and

MA

High-Temperature Hydrodynamic Phenomena, Mineola, Now York: Dover

AC

CE P

TE

D

Publications Inc. 2002, page 385.

26

ACCEPTED MANUSCRIPT Highlights of the manuscript entitled “Morphology and characteristics of laser-induced aluminum plasma in argon and in air: a comparative study” by Xueshi Bai

IP

T

et al.

- Comparative study of the morphologies of the plasmas induced in argon and air

SC R

ambients.

- Influence of the ambient gas on the characteristics of the plasma, electron density and temperature.

NU

- Consequence of the different laser-supported absorption waves on the plasma expansion behavior.

AC

CE P

TE

D

MA

- Correlation between the plasma expansion and the crater left on the target.

27